ITG modes with finite-β effects and the upper ηi stability regime

Abstract

It is shown that the toroidal ion temperature gradient (ITG) mode has an upper η i (higher η i ) stability regime for experimentally relevant parameter values in addition to the lower regime with the stability threshold (η ilow ) at η i around one (η i = L n /L T i where L n and L T i are the characteristic lengths for the density and ion temperature gradients). The ITG mode is studied with a focus on the upper η i stability regime and the β dependence (β = plasma pressure/magnetic pressure). The results of a gyrokinetic and a two-fluid model as well as a semilocal approximation are compared. It is shown that the upper stability threshold (η iup ) is very sensitive to and considerably reduced by finite-β effects. It is also sensitive to finite Larmor radius (FLR) effects and to ε n = 2L n /L B (L B is the characteristic length for the toroidal magnetic field gradient). Predictions and comparisons are made with data from a joint European torus (JET) optimized shear discharge and a JET hot ion H-mode (high-performance mode) discharge.

Secondary instability of electromagnetic ion-temperature-gradient modes for zonal flow generation

“…It is known that the linear ITG mode growth rate is reduced by electromagnetic fluctuations, resulting in a favourable scaling of confinement with β in QL theories [8][9][10][11] . Less studied is the role of electromagnetic effects for the generation of zonal flows.…”

Effects of parallel ion motion on electromagnetic toroidal ion temperature gradient modes in a fluid model

“…(10)- (13) in Ref. 10 identically, the ion density response then being expressed in terms of the electrostatic potential / only. Electromagnetic effects, accounted for by the A k dependence, then enter via the quasineutrality condition and via the Ampere law.…”

“…Assuming an axisymmetric toroidal (Tokamak) geometry with a low beta equilibrium with circular flux surfaces and a high mode number approximation, we may use the ballooning mode formalism where the spatial perturbation is assumed to be of the form 10,34 U ¼ U n ðr; hÞe …”

“…[1][2][3][4][5][6][7][8][9] Although a kinetic description of the plasma is more detailed than a fluid description, the latter may give simpler and more explicit results. [10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] Also the access of computing power is in general not crucial with fluid codes. They normally may be run for times longer than the transport time scale contrary to kinetic codes.…”

“…[26][27][28]. In this contribution, however, a semilocal mode approximation 10,35 has been used in order to avoid a system of equations higher than second order.…”